exif_service

A C++ microservice used to extract location and other forensic information from JPEG images.

How to run

The application needs to be compiled and linked to generate a binary for target platform. It can be run on host machine as well as in a container.

Features:

1. Written in C++ with no library dependencies.
2. Simple stateless microservice which can be hosted on cloud or on-premise.
3. JSON logging support with ISO timestamp embedding.
4. Health endpoint to monitor the service's availability.
5. Native tool, which can be used as a CLI command to get JPEG EXIF data without running the server.

Run on host machine :

Any x86 Linux host can easily run this binary. Follow the instructions below : GNU G++ and make are enough to build this on any linux distro.

1. Generate the binary : We have provided a makefile to ease build process.

make web

This will generate the binary at build/server

2. Run the binary : We can run the binary using following command :

$./build/server
Listening on http://0.0.0.0:8888

By default, the logs will be redirected to stdout. Pass the log-file as argument to enable file-logging.

./build/server $PATH_TO_LOG_FILE/log.txt

This will run the server.

Build a docker image :

To reduce the size of docker image, we have used alpine-linux as base with minimal required dependencies.

1. Build the container image: Build the image using command below :

docker build . -t exif_service

2. Spin up the container : We can use docker run the container, the log directory is expected to be mounted at /logs.

docker run --rm -ti --net=host -v ${LOG_FILE_PATH}:/logs exif_service

Using the REST Api:

The server provides two rest endpoints :

1. Check health of the service : This endpoint is just like ping.

curl http://localhost:8888/health

2. Run exif extraction: The endpoint jpeg takes multipart/form-data and produces JSON output.

curl -X POST -F 'image=@{IMAGE_ABS_PATH}/image.jpg' -H Content-Type:multipart/form-data" http://localhost:8888/jpeg | json_pp

The output is a list of complete image information as per EXIF standard.

{
   "focal_plane_xres" : 0,
   "35m_focal_length" : 0,
   "image_original_date_time" : "",
   "image_width" : 0,
   "expousre_bias" : 0,
   "image_digitized_date_time" : "",
   "exposure_time" : 0,
   "f_stop" : 0,
   "software" : "",
   "subject_distance" : 0,
   "iso_speed" : 0,
   "lens_f_stop_min" : 0,
   "bits_per_sample" : 0,
   "image_date_time" : "",
   "metering_mode" : 0,
   "flash_mode" : 0,
   "gps_precision" : 0,
   "gps_alatitude" : 0,
   "subsecond_time" : "",
   "lens_f_stop_max" : 0,
   "flash_returned_light" : 0,
   "image_orientation" : 0,
   "exposure_program" : 0,
   "location" : {
      "lat" : {
         "minutes" : 0,
         "degree" : 0,
         "direction" : 63,
         "seconds" : 0,
         "lat" : 0
      },
      "lon" : {
         "degree" : 0,
         "minutes" : 0,
         "direction" : 63,
         "lon" : 0,
         "seconds" : 0
      }
   },
   "camera_make" : "",
   "lens_min_focal_length" : 0,
   "focal_plane_yres" : 0,
   "image_copyright" : "",
   "lens_make" : "",
   "camera_model" : "",
   "image_description" : "",
   "lens_focal_length" : 0,
   "lens_max_focal_length" : 0,
   "lens_model" : "",
   "image_height" : 0,
   "flash_used" : 0
}

The values are purposefully emptied/zeroed to protect sensitive information.

Acknowledgements :

1. EasyEXIF
2. cpp-httplib
3. json

Contributing :

Contribute to this repository by opening issues or directly making a PR!!